# Background required for understanding Robust PCA and Low-Rank Sparse Decomposition

My current knowledge is Linear Algebra, basics of statistics and Machine Learning (Andrew Ng's ML Coursera).

I have a very good understanding of classic PCA and I know how to implement it in python or Matlab. However, every time I watch a video or read a paper about Robust PCA, my mind just goes blank due to many equations, notations and names I don't understand.

What background should I have in order to comfortably understand Robust PCA and low-rank sparse decomposition?

• Welcome to our site! If you're interested in background reading you might need to do, consider adding the "references" tag – Silverfish Oct 25 '16 at 23:49
• Is there a particular paper you can point to that is giving you trouble? I am not familiar with the term "Robust PCA" (and the Wikipedia page is pretty vague), while "low rank sparse decomposition" is non-unique, and could describe several methods. One concept you do not mention that may be relevant is $L_1$ optimization, which is common in both robust statistics and sparse approximation. – GeoMatt22 Oct 26 '16 at 1:48
• @GeoMatt22 I'm particularly interested in understanding Emmanuel's paper [link](statweb.stanford.edu/~candes/papers/RobustPCA.pdf) – Anwar Oct 26 '16 at 8:23